c accepted manuscript not copyedited - university of surreyepubs.surrey.ac.uk/811908/1/jeecs-16-1019...

1

Carbonate dynamics and opportunities with low temperature, AEM-based electrochemical 1

CO2 separators 2

3

William A. Rigdon,a,b Travis J. Omasta,a,b Connor Lewis,a,b Michael A. Hickner,c John R. 4

Varcoe,d Julie N. Renner,e Kathy E. Ayerse and William E. Mustaina,b,* 5

6

a Department of Chemical & Biomolecular Engineering, University of Connecticut, Storrs, CT. 7

b Center for Clean Energy Engineering, University of Connecticut, Storrs, CT. 8

c Department of Materials Science and Engineering, Pennsylvania State University, State 9

College, PA. 10

d Department of Chemistry, University of Surrey, Guildford, UK. 11

e Proton OnSite, Wallingford, CT. 12

* Corresponding Author: [email protected]

14

KEYWORDS: carbon dioxide (CO2), electrolysis, carbonate, bicarbonate, electrochemical 15

separation (pump), anion exchange membrane (AEM) 16

17

AUTHOR INFORMATION 18

Corresponding Author 19

* email: [email protected]; phone: +1 860-486-2756; Address: 191 Auditorium Road, Unit20

3222. Storrs, CT 06269-3222 21

Journal of Electrochemical Energy Conversion and Storage. Received February 10, 2016; Accepted manuscript posted April 20, 2016. doi:10.1115/1.4033411 Copyright (c) 2016 by ASME

Accep

ted

Manus

crip

t Not

Cop

yedi

ted

Downloaded From: http://electrochemical.asmedigitalcollection.asme.org/ on 08/28/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

mailto:[email protected]

mailto:[email protected]

2

ABSTRACT 22

Fossil fuel power plants are responsible for a significant portion of anthropogenic atmospheric 23

carbon dioxide (CO2) and due to concerns over global climate change, finding solutions that 24

significantly reduce emissions at their source has become a vital concern. When oxygen (O2) is 25

reduced along with CO2 at the cathode of an anion exchange membrane (AEM) electrochemical 26

cell, carbonate and bicarbonate are formed which are transported through electrolyte by migration 27

from the cathode to the anode where they are oxidized back to CO2 and O2. This behavior makes 28

AEM-based devices scientifically interesting CO2 separation devices or “electrochemical CO2 29

pumps.” Electrochemical CO2 separation is a promising alternative to state-of-the-art solvent-30

based methods because the cells operate at low temperatures and scale with surface area, not 31

volume, suggesting that industrial electrochemical systems could be more compact than amine 32

sorption technologies. In this work, we investigate the impact of the CO2 separator cell potential 33

on the CO2 flux, carbonate transport mechanism and process costs. The applied electrical current 34

and CO2 flux showed a strong correlation that was both stable and reversible. The dominant anion 35

transport pathway, carbonate vs. bicarbonate, undergoes a shift from carbonate to mixed 36

carbonate/bicarbonate with increased potential. A preliminary techno-economic analysis shows 37

that despite the limitations of present cells, there is a clear pathway to meet the US DOE 2025 and 38

2035 targets for power plant retrofit CO2 capture systems through materials and systems-level 39

advances. 40

41

42


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


3

1. INTRODUCTION 43

The threat of global climate change has brought considerable attention to the need for a 44

reduction in anthropogenic CO2 emissions. Captured CO2 can be injected into existing geological 45

formations or oil reservoirs, which can dramatically increase the productivity of previously 46

depleted wells as well as allow the simultaneous, permanent, and safe storage of CO2.[1,2] The 47

U.S. Geological Survey recently reported that the U.S. alone has the potential for the storage of 48

3000 gigaton of CO2 (~550 years emissions).[3] The CO2 could also be used as an industrial 49

solvent, or even reduced chemically or biologically to fuels,[4,5] potentially creating new 50

economic opportunities and jobs. 51

The most significant sources for CO2 emissions are electric power plants, accounting for 52

around 35% of global emissions.[6,7] Unlike transportation emissions, a collection of millions of 53

small sources of CO2, there are only ca. 7300 electricity generation sites across the U.S., while 54

more than 50% of emissions come from the largest 250 plants.[8] These relatively few high 55

emission sites are prime targets for immediate action towards the reduction of CO2 released into 56

the Earth’s atmosphere. As energy demands are steadily rising and supply is met through the 57

combustion of fossil fuels, there is motivation to study methods that separate and concentrate CO2 58

from power plant flue gas. 59

Chemical sorption is widely viewed as the state-of-the-art technology for scrubbing CO2 from 60

flue gas. A thermo-chemical amine (e.g. monoethanolamine) solvent-based absorption process is 61

typically used for capture through chemical reduction. Recovery is achieved through thermal 62

oxidation, releasing a concentrated CO2 gas stream. There are two major disadvantages to using 63

chemical sorption for CO2 capture at larges scales: 1) The necessity for large amounts of sorbent 64

material that scales with the amount of CO2 captured, rapidly increasing the system size and cost 65


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


4

with scale-up. Theoretically, monoethanolamine requires two molecules for every CO2 molecule 66

removed.[9] Additionally, flue gas CO2 is dilute (12-14%), introducing mass transport issues into 67

the system. To compensate for this, some studies have shown that the amine:CO2 ratio can be as 68

high as 9;[10] and 2) The heating requirement to regenerate the amine.[11] A recent report 69

projected that adding an amine sorption CO2 system to a new pulverized coal power plant would 70

increase the cost of electricity by 80% and de-rate the plant’s net generating capacity by 71

approximately 30%;[1] other studies show that the energy penalty may be as high as 45%.[6,10,12] 72

Also, the steam temperature needed for efficient CO2 desorption in the scrubber thermally 73

degrades the amine over time to corrosive byproducts that cause vessel corrosion and additional 74

costs are incurred because of the need to replace the capture solvent. 75

Though amine-based CO2 capture processes have been in continuous development since the 76

1930’s, and there are small power and industrial plants that have used this process to recycle CO2 77

as a solvent,[12] the cost-benefit analysis is quite different for larger CO2 capture and storage 78

compared to these existing markets. The U.S. Department of Energy (DOE) has set a 2020-2025 79

capture target of $45/tonne of CO2 for retrofit coal-fired power plants and even more stringent cost 80

targets for new plants.[1] Existing sorption systems have an average CO2 capture cost of $61/tonne 81

of CO2 (assuming a generation revenue loss of $0.075 USD/kWh – the peak industrial rate in 82

2014[13] – and a capital cost amortization of 6 years with a $6 USD/tonne of CO2 operational 83

cost[12]) and it will still be decades before the existing targets are met.[1] The excessive energy 84

requirements and high cost of state-of-the-art amine-based CO2 capture systems are the same 85

reasons that sorption technology was considered unacceptable at the power plant scale 25 years 86

ago.[14] Therefore, there is an urgent need for new technologies that approach CO2 capture from 87


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


5

a fresh perspective in order to meet these challenging cost targets with a reasonable development 88

time. 89

Electrochemical processes are promising alternatives for the separation, concentration, and 90

compression of CO2 since they are not bound by thermochemical cycles and so their theoretical 91

energy requirements will always be lower [15–17] [REFS. High temperature electrochemical cells 92

based on ceramic and molten carbonate electrolytes have been considered [18–21], but these 93

systems require high heat input and suffer dynamic (operation) instabilities in addition to the 94

electrical cost, making low temperature systems more suitable for large-scale CO2 separation. One 95

low temperature electrochemical approach that has been recently explored is the sorption of CO2 96

by electrochemically reduced disulfides (thiolates) in tetraalkyl phosphonium/ammonium ionic 97

liquid (IL) leading to the formation of thiocarbonates.[22–24] In a secondary step, the 98

thiocarbonates are chemically transformed back to the disulfides and high purity CO2 is released. 99

This process has some very attractive features, including a mechanism naturally requiring only 1 100

electron per CO2 molecule and there is also no need to generate steam for high quality heat; both 101

act to reduce the energy required for CO2 capture and highlight the technology’s potential. 102

However, there are a few potential drawbacks of this approach including: 1) likely high operating 103

voltage; 2) low gas mass transport into IL; 3) low ionic conductivity for IL; and 4) cost and 104

scalability of IL. 105

Low-temperature anion exchange membrane (AEM) electrochemical CO2 separators are a 106

relatively unexplored, yet promising, technology for low energy, low cost CO2 separation from 107

power plant flue gas. AEM-based CO2 separators are in their infancy with limited previous work 108

reported. The very earliest work utilized a porous membrane soaked in aqueous bicarbonate 109

electrolyte with nickel mesh electrodes.[25–27] These devices were run at very high operating 110


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


6

voltages, which led to energy requirements for CO2 separation that would be 200% of the generated 111

power plant energy in some cases. Another serious problem with such high operating voltages is 112

water electrolysis and unwanted corrosion reactions, resulting in low faradaic efficiencies. More 113

recently, Kitchin et al. successfully operated an electrochemical CO2 device with nickel catalysts 114

at 1.2 V at room temperature.[28,29] Their device had a significantly improved energy footprint 115

over previous electrochemical devices, requiring only 78% of the power plant output to separate 116

CO2 (approximately two times that of amine sorption) despite the fact that they used low surface 117

area catalysts and lower performing anionic ionomer compared to the existing state-of-the-art. 118

Additionally, their work focused primarily on the anode catalyst, with minimal consideration for 119

the cathode catalyst, electrode structure, and electrolyte; probably the most important components 120

which control the reaction selectivity and anion transport resistance of the cell. Additionally, their 121

work did not capture the potential dependence of the reaction selectivity and CO2 separation 122

performance, and hence very little is currently known about the system dynamics. Therefore, 123

several scientific and engineering questions remain and significant opportunities are available for 124

rapid and transformational innovations in component and cell design, operation, system costs, and 125

performance. 126

In this study, the potential voltage dependent dynamics of CO2 separation are explored through 127

coupled (simultaneous) current density and CO2 effluent concentration measurements. Several 128

important characteristics are assessed including: the operating current, the relationship between 129

current and CO2 exchange rate, cell reversibility and performance hysteresis, the dominant anion 130

transport pathway, electrical cost, and the impact of operation on the thermodynamic efficiency of 131

a typical coal fired power plant. We also report a preliminary techno-economic analysis that 132


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


7

explores both the energy and capital costs of the system to show the promise of AEM-based 133

electrochemical separators to meet cost targets at the 500 MW power plant scale. 134

135

2. THEORY & OPERATION OF AEM-BASED CO2 SEPARATORS 136

An illustration showing the operating principle of an AEM-based CO2 separation cell is 137

presented in Figure 1. 138

139

Figure 1. Operating principles of an AEM electrolyte electrochemical CO2 separator. 140

141

The exhaust from a coal-fired or natural gas-fired power plant is fed to the cathode where the CO2 142

and O2 are separated from the incoming flue gas via electrocatalytic reduction to carbonate 143

(CO32-) and bicarbonate (HCO3

-) anions. An indirect carbonate path requires O2 reduction as a 144

primary step (Equation 1) – the reaction then proceeds through non-electrochemical pathways 145

initially yielding HCO3- (Equation 2) followed by CO3

2- (Equation 3). 146

A direct carbonate path is also possible for CO32- generation (Equation 4), where bicarbonate 147

can be subsequently produced via the reverse of the reaction presented in Equation 3. 148


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


8

O2 + 2H2O + 4 e- → 4OH- (1) 149

2OH- + 2CO2 → 2 HCO3- (2) 150

2OH- + 2HCO3- ↔ 2CO3

2- + 2H2O (3) 151

O2 + 2CO2 + 4 e- → 2CO32- (4) 152

The CO32-/HCO3

- anions are transported across an AEM to the anode where they are electrolyzed 153

back to CO2 and O2. The implications of Equations 1-4 are interesting, and critical to understanding 154

the operation of the device. No matter which reaction(s) dominate, four electrons are required per 155

mol of oxygen reduced. These electrons can be balanced by either carbonate or bicarbonate 156

transport through the electrolyte. If bicarbonate is the dominant anion, four CO2 molecules are 157

transported through the membrane; another way to consider this is that only 1 electron is required 158

per separated CO2 (n = 1). If carbonate is the dominant anion, its divalence leads to only 2 CO2 159

molecules being separated per reduced oxygen, or that 2 electrons are required per separated CO2 160

(n = 2). This has practical implications to these devices, which will be discussed further in Section 161

4. 162

In practical engineered systems, the anode carrier gas will be defined by the final application. 163

In the context of a power plant, the carrier gas for the CO2/O2 stream would most likely be methane, 164

which would be oxy-combusted to yield high purity CO2 after condensing out the water.[28] Of 165

course, the oxy-combustor and condenser will add capital cost to the system, and the condenser 166

will require electricity to operate; however this can be offset by the energy produced through oxy-167

combustion. It should also be noted that the water balance and management in the AEM-based 168

CO2 separator will need to be considered in larger engineered system. Due to the relative infancy 169

of these devices, the costs and credits associated with the oxy-combustor, and condenser and 170

recycle will not be elaborated in this work. The carrier gas used here was high purity N2. 171


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


9

Carbonates (and CO2) have long been considered, and treated as, poisons in low temperature 172

aqueous alkaline systems (i.e. alkaline fuel cells containing aqueous KOH electrolytes) due to the 173

low solubility of alkali metal carbonate salts in free water. However, AEMs do not form these 174

insoluble compounds because the carbonate anions are associated with stationary (already solid 175

state) cationic groups fixed to the polymer backbone.[30] As a result, carbonate-based AEM 176

systems have become an active research topic as interest in alkaline membrane electrochemistry 177

and electrochemical devices has drastically increased in the last decade.[30–33] Some research 178

has even targeted the specific utilization of carbonates through the study of new catalysts and 179

understanding of fundamental anion transport/exchange mechanisms.[5,34–41] Use of a 180

(bi)carbonate conducting polymer electrolyte cell has several important implications for 181

commercial applications. Although the development of these cells needs to be advanced, 182

improvements could result in a range of exciting new technologies. Separation of CO2 from flue 183

gas exhausted at power plants is just one promising possibility since capture and compression of 184

CO2 is valuable to many other applications. A low temperature, low cost anion conducting fuel 185

cell that operates on air and resists carbonation can become a real possibility.[42] Furthermore, the 186

conversion (partial oxidation) of hydrocarbon fuels to value added products in the anode of an 187

electrochemical carbonate cell may be a further possibility.[39] Notably, CO2 reduction and 188

carbonate anion-exchange yield a large range of new opportunities at low (potentially near room) 189

temperatures. 190

191

3. MATERIALS AND METHODS 192

An AEM-based CO2 separator cell was assembled with commercial Pt cathode and anode 193

catalysts, a poly(2,6-dimethyl-1,4-phenylene oxide) (PPO) quaternary ammonium (QA) 194


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


10

AEM[31], a radiation-grafted QA anion-exchange ionomer (AEI),[43] and operated at 50 oC with 195

no back pressure. 196

197

3.1 Electrode preparation 198

50 wt% platinum on Vulcan XC-72R (BASF) electrocatalysts were used at both the anode and 199

cathode. AEI powders (ion exchange capacity, IEC, = 1.24 mmol g-1) were prepared from 200

radiation-grafted ETFE-powder and contained benzyltrimethylammonium functionality.[43] The 201

AEI was hand-ground (milled) with a mortar and pestle for 10 min with dry electrocatalyst and the 202

AEI comprised 15% of total catalyst layer mass. The solids were wetted with 2-3 mL DI water 203

before suspension in 10 mL of isopropanol solvent using ultrasonic mixing for 30 min to form an 204

electrode ink. Ink suspensions were spray deposited with an air brush onto Toray PTFE-treated 205

carbon paper (TGP-H-030) to achieve a 0.5 mgPt/cm2 loading; this was then cut into 5 cm2 square 206

gas diffusion electrodes (GDE). 207

208

3.2 Polymer anion exchange membrane 209

The synthesis as well as the physical and electrochemical characterization of the PPO 210

membrane used in this work was extensively described previously [31]. Briefly, PPO with ca. 211

40% degree of bromination (40 % of the repeat units were brominated) was reacted with 212

trimethylamine (Me3N) and then cast into an AEM. The average thickness of the dry AEM was 213

57 ±3 μm and the IEC = 2.20 mmol g-1 (calculated in the Br- form). Before cell construction, the 214

AEM was soaked in aqueous Na2CO3 (1 M) for 2 h and then thoroughly rinsed in DI water (to 215

remove the excess Na2CO3 species). 216

217


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


11

3.3 Cell construction and operation 218

A symmetric single cell was assembled using Fuel Cell Technologies (Albuquerque, NM) 219

hardware with carbon graphite plates containing single serpentine flow channels and gold plated 220

current collectors. Membrane electrode assemblies were fabricated in the cell hardware at room 221

temperature by sandwiching the AEM between the GDE-supported anode and cathode electrodes. 222

The GDEs were bordered with 127 μm thick Tefzel ETFE gaskets, resulting in ca. 30% 223

compressive pinch. Once aligned, the cell was sealed using 7 N·m torque on the eight cell sealing 224

bolts. A Scribner 850e test station controlled the temperature of the cell (50 °C) and the gas flows 225

(100 % relative humidity, 0.2 standard L min-1). The cathode gas supply was equivalent volumes 226

of CO2 and O2 (with the exception of O2-only control experiments), while high purity N2 was 227

supplied to the anode. It should be noted that tests were performed several times on multiple cells 228

and the cell performance and behavior was highly repeatable. A representative data set from a 229

single cell is presented. 230

231

3.4 Electrochemical cell tests 232

An Autolab PGSTAT302N was used to control the applied cell potential. Dynamic test were 233

run using linear sweep voltammetry (LSV). LSVs were collected from 0.0 to +1.0 V at 5 mV s-1 234

scan rate. A series of chronoamperometry (CA) steps were also performed from +0.1 V to +1.5 V 235

and then back to +0.1 V with each step held for 30 min. Electrochemical impedance spectroscopy 236

(EIS) experiments were used to determine the high frequency resistance (10 mV perturbations and 237

frequency sweep from 50 → 0.5 kHz). 238

239

240


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


12

3.5 Carbon dioxide measurements 241

The CO2 gas evolved from the anode was monitored continuously using a PP Systems SBA-5 242

gas analyzer and accompanying software, which utilizes a highly CO2 selective non-dispersive 243

infrared (NDIR) spectroscopy technique to measure the CO2 concentration. The baseline CO2 244

crossover in the cell was determined prior to each experiment and was subtracted from the raw 245

data. The anode exhaust was first passed through an ice water chilled glass condensation bulb to 246

prevent liquid from entering the detection chamber. The sweep gas flow rate was measured using 247

an Agilent Digital Flowmeter Optiflow 650 and was used in estimating the molar evolution rates 248

of CO2. The high N2 anode carrier gas flowrate (200 mL/min), was used in order to limit the CO2 249

content of the anode < 1000 ppm in order to stay within the NDIR calibrated range and minimize 250

response time. Therefore, the detected and reported CO2 content of the anode is not a limiting 251

value, it is purposefully, and artificially, low. 252

253

4. RESULTS & DISCUSSION 254

To verify that CO2 was removed from the cathode feed due to the electrical stimulus and not 255

diffusion, control experiments were initially performed. First, humidified O2 was fed to the cathode 256

and then, in a second experiment, the cathode was supplied with a feed containing an equivalent 257

volume mixture of O2 and CO2. Both cells were polarized by LSV up to a cell potential of 1.0 V 258

and coupled cell current/CO2 anode emission measurements were recorded (Figure 2). Notably, 259

the polarization curve showed a decrease in performance when operated with only O2 and the CO2 260

evolved at the anode was significantly less than the experiment with 50% CO2. The amount of 261

evolved CO2 at the anode in the O2-only cathode supply control experiment was non-zero and has 262

several possible origins. One source may be de-carbonation of the AEM when the OH- anions 263


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


13

produced at the cathode displace CO32-/HCO3

- anions persisting in the polymer electrolyte. 264

However, carbon support and GDL material corrosion at higher anode potentials is also 265

possible.[44,45] Next, the cell was allowed to rest at the open circuit voltage with the 50:50 CO2/O2 266

cathode supply and the CO2 crossover was measured as 1.6 × 10-9 mol s-1. The amount of CO2 that 267

could be transported by electro-osmotic drag was overestimated with the assumption of a relatively 268

high drag coefficient of 4 mol of CO2-saturated H2O (Henry’s Law) per e-, which accounted for 269

less than 2% of the value measured in the effluent at +1.0 V. Each of these secondary CO2 270

transport mechanisms (diffusion and drag) were unable to account for the CO2 observed in the 271

anode effluent during cell testing. This confirms that the driving force for the CO2 separation was 272

the redox chemistry occurring at the electrodes. 273

274

Figure 2. Comparison of applied current and CO2 evolution at the anode with 0 %vol and 50 275

%vol CO2 in the O2-based cathode feed. The N2 flowrate was set at 200mL/min to purposefully 276

reduce the CO2 concentration to < 1000 ppm. 277

Since the electrochemistry is the driving force for the CO2 pumping action in this cell, it is 278

important to understand how the cell potential impacts the electrochemical selectivity and 279

effectiveness of the system. Steady-state observations showed a strong correlation between applied 280

cell current and CO2 evolution at the anode (Figure 3a). To measure the steady-state performance, 281

a staircase was used where each 0.1 V step was held constant for 30 min while polarizing the cell 282


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


14

from +0.1 V → +1.5 V → +0.1 V. After each step change, the current and CO2 crossover had 283

sufficient time to stabilize, and the final data point from each step was taken to be the steady-state 284

current. The data comparing the steady-state current and CO2 evolution rate with respect to cell 285

potential are presented in Figure 3b. Tests were performed several times and the cell performance 286

and behavior was highly repeatable. 287

288

289

Figure 3. (a) Correlation of the applied cell current (black) and measured CO2 evolution rate at 290

the anode (red). (b) Current and CO2 evolution rate vs. cell potential (forward and backward 291

polarizations are shown). 292

293

The CO2 evolution reversibly tracked the cell potential as shown in the overlapping anodic and 294

cathodic polarization curves of Figure 3b, particularly at higher currents. There was minimal 295

(a)

(b)


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


15

hysteresis from the responsive system. The AEM and AEI used in this cell demonstrated good 296

mechanical integrity and visual inspection of the membrane electrode assembly before and after 297

testing suggested low dimensional swelling under full humidification. The high frequency area 298

resistance (from EIS) was somewhat large in the representative data set, 55 Ω m, a value that needs 299

further evaluation and improvement for a commercially-viable electrochemical reactor. 300

As the cell potential is increased, the cathode potential shifts negative and the relative driving 301

forces for the reactions in Equations 1 – 4 change. The dynamic driving force with voltage means 302

that the mechanism for anion formation and transport can change. The dominant anion transport 303

(CO32- vs. HCO3

-) mechanism was estimated from the number of electrons required to move each 304

CO2 molecule (n) from the feed stream (cathode) to the exhaust stream (anode). Equation 5 shows 305

the relationship between n and the cell operating current (i), Faraday’s constant (F), the 306

concentration of CO2 in the anode stream in ppm (CCO2) and the molar flowrate of the sweep gas 307

in mol s-1 (MG) . 308

𝑛 = 106∗𝑖

𝐹∗𝐶𝐶𝑂2∗𝑀𝐺 (5) 309

The voltage dependence for the number of e- required to transport each CO2 molecule is plotted 310

in Figure 4. At very low currents (cell potential < 0.7 V), the amount of CO2 measured at the 311

anode was close to the NDIR detection limit, and n was highly sensitive to slight variations in 312

signal, leading to an artificial increase in n. Therefore, only data collected at cell potentials > 0.7 313

V were used to elucidate the potential dependence of the CO32-/HCO3

- mechanism (represented in 314

Figure 4). At 0.7 V, n was slightly above 2, affirming CO32- conduction as the dominant anion 315

transport pathway at low cell potentials. As the cell potential increased, there was a clear transition 316

from CO32- to a near equal balance of CO3

2-/HCO3-, where n stabilized around a value of 1.5. 317


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


16

318

Figure 4. The number of electrons (blue) required to separate each CO2 molecule coupled with 319

the separation energy and electrical costs (green) as a function of the cell potential. 320

321

4.1 Energy requirements 322

The typical output of a coal-fired power plant normalized to its CO2 emissions is approximately 323

1.1 MWh per tonne of CO2. In AEM-based CO2 separators, the energy requirement (ES) for 324

separation is controlled primarily by the cell potential (V) and n as defined by Equation 6. 325

𝐸𝑆 (𝑀𝑊ℎ

𝑡𝑜𝑛𝑛𝑒𝐶𝑂2) =

𝑛𝐹|𝑉|

44.01∗3600 (6) 326

Therefore, the anode and cathode catalysts need to have low activation overpotentials, and the 327

AEM must have high anion conductivity to minimize the cell potential. In addition, the cathode 328

catalyst must provide the appropriate reaction selectivity, since the amount of energy required 329

doubles when CO32- (n = 2) vs. HCO3

- (n = 1) production is favored at the same cell potential. On 330

the practical size, HCO3- is preferred mechanistically because of its much lower energy cost; 331

however, bicarbonate operation does have one very important tradeoff: HCO3- has a lower intrinsic 332


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


17

mobility than CO32- due to its larger hydration radius, which will lead to higher electrolyte (AEM) 333

ionic resistivity thus higher cell voltage. The PPO membrane in this work was selected because 334

of its high carbonate conductivity, relatively low water uptake and good mechanical strength [31]. 335

However, chemical and mechanical durability are not the most important factors when choosing 336

an AEM for this application since AEMs are intrinsically more stable in carbonate form than 337

hydroxide form [30,46]. 338

One distinct advantage of AEM-based CO2 separation vs. chemical sorption is that the 339

thermodynamic minimum energy requirement for CO2 separation is 80% lower since 340

electrochemical systems are not bound by thermochemical cycles. The minimum energy 341

requirement for chemical sorption is approximately 11% of the power plant rating (~ 0.12 MWh 342

per tonne CO2) based on the heating requirement to produce steam and release CO2 from the amine 343

sorbent.[12] In contrast, the minimum for the AEM separator studied in this work is only 0.029 344

MWh per tonne CO2 based on the Nernst equation (Equation 7), which is 2.6% of the power plant 345

rating if the device is operated at 50 oC (assuming an exclusive bicarbonate pathway). 346

𝑉𝑇 =𝑅𝑇

𝑛𝐹𝑙𝑛 (

𝑃𝐶𝑂2,𝑠𝑒𝑝

𝑃𝐶𝑂2,𝑓𝑙𝑢𝑒𝑔𝑎𝑠) (7) 347

where VT is the thermodynamic cell voltage, R is the ideal gas constant, T is the temperature (K), 348

PCO2,sep is the partial pressure of CO2 in the anode exhaust, and PCO2,fluegas is the partial pressure of 349

CO2 in the cathode feed. Therefore, electrochemical AEM-based CO2 separators have the potential 350

for energy requirements that are not only less than the state-of-the-art, but are impossible to achieve 351

with amine sorption. 352

The electrochemical operating space for a CO2 separator is shown in Figure 5a. Operating at or 353

above the power plant generation energy (1.1 MWh per tonne CO2) is represented by the red region 354

in Figure 5a. The span of energy requirements for existing chemical sorption technologies are 355


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


18

shown in orange. To date, investigators have not been able to achieve energy requirements that 356

approach the thermodynamic limit for electrochemical-based CO2 separation, though this is to be 357

expected given the limited state of development of the technology; however, significant progress 358

has been made. 359

360

Figure 5. (a) Existing operating requirements for amine and electrochemical separation showing 361

how AEM-based electrochemical cell improvements can yield energy requirements below the 362

thermodynamic limit for chemical sorption; (b) influence of cell and stack improvements on the 363

cost of electrochemical CO2 separation, showing that AEM-based electrochemical cell 364

improvements can lead to very low costs (ca. ⅓ of chemical amine sorption). 365

366

The first AEM cells operated around 2.5 V, which is an energy requirement of 1.57 MWh per 367

tonne CO2 if n = 1 (exclusive HCO3- pathway) or 3.14 MWh per tonne CO2 if n = 2 (exclusive 368

CO32- pathway); these represent 140 and 280% of the power plant energy to operate, respectively, 369

which are obviously far too high for practical application.[25–27] Landon and Kitchin were able 370

to reduce the energy requirement to 0.88 MWh per tonne CO2 (~78% power plant output) by 371

reducing the operating voltage to 1.2 V operating mostly on the HCO3- cycle (light blue square in 372

Figure 5a).[29] Using Equation 5, the energy requirement for CO2 separation was calculated for 373

the staircase experiments as a function of the cell potential (Figure 4). For this representative data 374


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


19

set, the lowest energy requirement was 0.93 MWh per tonne CO2; however, our best performing 375

cell had a lower n = 1.47 at a cell potential of 0.9 V, yielding a CO2 separations energy of 0.80 376

MWh per tonne CO2 (72% of power plant output). Though this number will need to be improved, 377

it represents the lowest energy requirement reported in the literature to date (dark blue diamond in 378

Figure 5a). 379

In order to achieve AEM-based cells that approach the thermodynamic limit, researchers must 380

not only make material advances to improve the cathode selectivity for bicarbonate, but also reduce 381

the electrode overpotentials and membrane resistance. They should eliminate gaps in the scientific 382

and operational knowledge by understanding the voltage dependence of the CO2 separation 383

dynamics, the relationship between current and CO2 exchange rate, cell reversibility and stability, 384

as well as the anion transport and redox mechanisms. These were all explored in this work; 385

however, there is considerably more work that needs to be done to determine the influence of 386

temperature, gas composition, impurities (i.e. NH3, H2S are well known Pt poisons that are present 387

in flue gas), as well as cell construction and operation variables on its performance. These will be 388

the focus of our future work. 389

390

4.2 Preliminary Unit Operations Considerations and Cost 391

There are three primary cost drivers from a unit perspective: lost electrical generation, capital 392

investment and amortization, and plant operation and maintenance (O&M). In amine scrubbing 393

systems, the lost electrical generation cost stems from thermal de-rating of the power plant, while 394

in the AEM-based system, electricity is internally rerouted and cannot be sold. Equation 8 395

calculates the electrical cost per tonne of CO2 emitted from the power plant (results plotted in 396

Figure 4). 397


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


20

$

𝑡𝑜𝑛𝑛𝑒𝐶𝑂2= 1000 ∗ 𝐸𝑆 ∗ $0.075/𝑘𝑊ℎ (8) 398

For the representative data set in Figure 4, where the lowest energy requirement was 0.93 MWh 399

per tonne of CO2, the lost electrical generation cost would be $70 per tonne of CO2; however, the 400

cell with the lowest measured e-/CO2 ratio of n = 1.47 (at 0.9 V) would have an electrical cost of 401

only $60 per tonne of CO2. Though this number is reasonably close to the cost target for the U.S. 402

DOE 2025 target for a retrofit coal-fired power plant, it is only part of the picture since the capital 403

and O&M costs must also be considered. 404

From a capital costs perspective, the primary drivers are the system materials (what is used for 405

the cell hardware, catalysts, membranes and other components) and the operating current density. 406

The flow of CO2 out of a large coal-fired powerplant is very large; e.g. the rate from a typical 500 407

MW coal-fired power plant is more than 7.5 tonne min-1. In order to operate at this scale, thousands 408

of parallel CO2 separator stacks would be needed to achieve complete separation and the number 409

scales almost linearly with the operating current density. Based on the proprietary costs and 410

estimates for existing Proton OnSite commercial stacks with their standard materials for the flow 411

fields and cell separators, a 2 mA cm-2 operating current (consistent with Figures 3 and 4), and the 412

same catalysts and AEM/AEI used in this work, and amortizing the capital cost over 6 years, it 413

was estimated that the existing CO2 pump capital cost would be approximately $2300 per tonne of 414

CO2. This is currently a very large cost, but not unexpected for an immature technology; this 415

highlights the need to improve the cell performance and lower the cost of the materials of 416

construction, which will be discussed in detail below. 417

Contributing to the excessive cost are three materials and systems-level properties that must 418

all be improved in order to exceed the U.S. DOE 2025 cost targets: 419


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


21

1) Operating current density, which dictates the total system size and right now is the primary 420

driver of the capital cost, should be increased from 2 mA cm-2 to at least 200 mA cm-2 by improving 421

catalyst and AEM chemistry and structure, as well as improved electrode engineering. We can 422

foresee the possibility since similar approaches have raised the operating current density of OH--423

based AEM systems from < 100 to > 1000 mA cm-2 in the past five years, while simultaneously 424

reducing the cell overpotentials.[30] One important consideration is that high current devices may 425

have a higher portion of ions transporting through the system in OH- form, which would, in effect, 426

reduce the faradaic efficiency of the device. Next-generation cells must maintain efficiencies, 427

which can be achievable through carbonate-selective catalysts or membranes with high CO2 428

permeability. 429

2) Replacing high cost materials, including titanium cell components with stainless steel, and 430

platinum catalysts with non-noble metal catalysts[47] is facilitated in this concept because of the 431

low corrosion CO32-/HCO3

- (mildly alkaline) environments. 432

3) Reducing the cell operating voltage below 0.5 V (about 5 × the voltage required for existing 433

H2 pumping cells operating at 200 mA cm-2) through innovations in the catalyst, AEM and AEI, 434

while simultaneously reducing the e-/CO2 ratio below n = 1.2. Possible metrics include reduction 435

of the cathode overpotential to < 0.3 V, anode overpotential < 0.2 V, and the combined AEM and 436

contact overpotential < 0.025 V. 437

The projected cost reduction of an AEM-based electrochemical CO2 separator with many of the 438

innovations discussed above is shown in Figure 5b. For all cases, a constant $6 per tonne CO2 439

O&M cost was assumed, which is consistent with sorption technology. Increasing the cell current 440

density from the existing 2 mA cm-2 to 200 mA cm-2 reduces the capital cost to a reasonable value 441

of $23 per tonne CO2, on par with the capital cost for a typical amine sorption system. It should 442


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


22

also be noted that increasing the AEM separator current density to 200 mA cm-2 yields a system 443

size of approximately 750 m3, which is around ½ of the volume of the absorber/stripper/boiler that 444

it would replace from the chemical sorption system. Further cost savings by transitioning to 445

stainless steel stack components reduces the capital cost to around $12 per tonne CO2. 446

Transitioning from Pt catalysts with moderate HCO3- selectivity to high selectivity (n = 1.2), non 447

Pt catalysts would reduce the capital cost further to $10 per tonne CO2 and simultaneously reduce 448

the electrical cost from $60/ per tonne CO2 to $48 per tonne CO2. Lowering the operating voltage 449

to 0.5 V reduces the electrical cost to $27 per tonne CO2. 450

When combined, these achievable innovations, which should be the R&D goals in the short 451

term, lead to a total cost for an AEM-based CO2 separations system of $43 per tonne CO2 – lower 452

than the U.S. DOE 2025 capture target of $45 per tonne CO2 for retrofit coal-fired power plants. 453

Future, long-term, innovations can be expected to further reduce the operating voltage to 0.25 V 454

(still more than 2× the value for H2 pumping), increase the HCO3- selectivity to yield n = 1 (e- per 455

CO2 molecule) and increase the operating current to 1 A cm-2 (consistent with the operating current 456

densities of state-of-the-art AEM-based fuel cells and electrolyzers). These innovations would 457

reduce the cost to around $22 per tonne CO2, which is even lower than the U.S. DOE 2035 cost 458

target for retrofit plants ($30 per tonne CO2).[1] Clearly, electrochemical CO2 pumping has a long 459

way to go in its development. In addition to materials advances an important consideration for 460

future study is the possible impact of common flue gas impurities at relevant concentrations on the 461

operating current density, reaction mechanism and capture efficiency, i.e. CO (20 ppm), 462

hydrocarbons (10 ppm), HCl (100 ppm), SO2 (800 ppm), and NOx (800 ppm). Also, the oxy-463

combustion and water management scheme must be considered, designed and analysed. However, 464

the calculations and experiments in this work highlight the exciting ultimate potential for AEM-465


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


23

based electrochemical CO2 separators as a low energy, low cost option for CO2 separation from 466

flue gas in the near future. 467

5. CONCLUSIONS 468

An AEM-based electrochemical CO2 separator was investigated and its future performance 469

and cost metrics discussed. In these devices, (bi)carbonate anions are produced through the 470

electrochemical reduction of O2 + CO2. The anions are transported from the cathode to the anode 471

through an anion-exchange membrane (AEM) where they are electrolyzed back to CO2 and O2. In 472

this work, the relationships between the cell potential and carbonate/bicarbonate selectivity were 473

explored, giving valuable insight to reactions occurring while offering a roadmap for future 474

investigations into improved cell design and construction materials. The cell current and CO2 475

separation through the AEM were closely correlated. Carbonate and bicarbonate both played a role 476

in anion exchange and transport through the cell, particularly at higher voltages. It was also found 477

that the energy required, electrical cost and capital cost offer a positive perspective on the possible 478

application of AEM-based electrochemical pumping technology for use in CO2 separation. The 479

AEM system reduces the thermodynamic barrier for CO2 separation by 80% compared to 480

conventional amine sorption process and there is a clear pathway to achieve systems costs that 481

meet U.S. DOE 2025 and 2035 targets for retrofit coal-fired power plants. Therefore, AEM-based 482

electrochemical CO2 separation systems are a promising area for future research with the potential 483

to have a high impact on the carbon capture and utilization landscape in the near future. 484

ACKNOWLEDGMENT 485

The cell assembly and performance work in Professor Mustain’s lab was supported by the U.S. 486

DOE Early Career Program through Award Number DE-SC0010531. Professor Hickner 487


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


24

acknowledges partial support of this work by the United States-Israel Binational Science 488

Foundation (BSF) through Energy Project No. 2011521 and the backing of industrial sponsors. 489

The anion-exchange ionomer powders were synthesized using funds from Professor Varcoe’s 490

EPSRC Fellowship (Grant EP/1004882/1): the raw IEC data for the AEI powders is available 491

without restriction [details at the University of Surrey publications repository at DOI: 492

dx.doi.org/10.15126/surreydata.00807830 (http://epubs.surrey.ac.uk/807830/). 493

ABBREVIATIONS 494

AEI, anion exchange ionomer; AEM, Anion Exchange Membrane; CA, chronoamperometry; EIS, 495

electrochemical impedance spectroscopy; ILs, Ionic Liquids; LSV, linear sweep voltammetry; 496

NDIR, non-dispersive infrared; PPO, poly-phenylene oxide; QA, quaternary ammonium; 497

REFERENCES 498

[1] 2013, Carbon Capture Technology Program Plan. 499

[2] Boot-Handford, M. E., Abanades, J. C., Anthony, E. J., Blunt, M. J., Brandani, S., Mac 500

Dowell, N., Fernández, J. R., Ferrari, M.-C., Gross, R., Hallett, J. P., Haszeldine, R. S., 501

Heptonstall, P., Lyngfelt, A., Makuch, Z., Mangano, E., Porter, R. T. J., Pourkashanian, M., 502

Rochelle, G. T., Shah, N., Yao, J. G., and Fennell, P. S., 2014, “Carbon capture and storage 503

update,” Energy Environ. Sci., 7, p. 130. 504

[3] 2013, National Assessment of Geologic Carbon Dioxide Storage Resources — Results. 505

[4] Dimitriou, I., García-Gutiérrez, P., Elder, R. H., Cuellar-Franca, R., Azapagic, A., and 506

Allen, R. W. K., 2015, “Carbon dioxide utilisation for production of transport fuels: process 507

and economic analysis,” Energy Environ. Sci., 8, pp. 1775–1789. 508

[5] Spinner, N. S., Vega, J. A., and Mustain, W. E., 2012, “Recent progress in the 509

electrochemical conversion and utilization of CO2,” Catal. Sci. Technol., 2(1), p. 19. 510


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


http://epubs.surrey.ac.uk/807830/

25

[6] Fisher, K. S., Street, S. A., Rochelle, G., and Figueroa, J. D., 2005, “Integrating MEA 511

regeneration with CO2 compression to reduce CO2 capture costs,” 4th Annual Conference 512

on Carbon Capture and Sequestration DOE/NETL, pp. 1–11. 513

[7] U.S. Environmental Protection Agency, 2015, “Greenhouse Gas Emissions” [Online]. 514

Available: http://www.epa.gov/climatechange/ghgemissions/. 515

[8] U.S. Energy Information Administration, 2013, “Independent Statistics and Analysis, 516

Electricity” [Online]. Available: 517

http://www.eia.gov/countries/analysisbriefs/Nigeria/nigeria.pdf. 518

[9] Yu, C., Huang, C., and Tan, C., 2012, “A Review of CO 2 Capture by Absorption and 519

Adsorption,” Aerosol Air Qual. Res., 12, pp. 745–769. 520

[10] Zhai, R., and Yang, Y., 2010, “MEA-Based CO2 Capture Technology and Its Application 521

in Power Plants,” Paths to Sustainable Energy, J. Nathwani, and A. Ng, eds., InTech, pp. 522

499–511. 523

[11] Ho, M. T., Allinson, G. W., and Wiley, D. E., 2008, “Reducing the Cost of CO2 Capture 524

from Flue Gases Using Pressure Swing Adsorption,” Ind. Eng. Chem. Res., 47, pp. 4883–525

4890. 526

[12] Rochelle, G. T., 2009, “Amine scrubbing for CO2 capture.,” Science, 325(5948), pp. 1652–527

1654. 528

[13] U.S. Energy Information Administration, 2015, “Independent Statistics & Analysis, 529

Electricity, Sales (consumption), revenue, prices & customers” [Online]. Available: 530

http://www.eia.gov/electricity/data.cfm#sales. 531

[14] Booras, G. S., and Smelser, S. C., 1991, “An engineering and economic evaluation of CO2 532

removal from fossil-fuel-fired power plants,” Energy, 16(11-12), pp. 1295–1305. 533


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


26

[15] White, C. M., Strazisar, B. R., Granite, E. J., Hoffman, J. S., and Pennline, H. W., 2003, 534

“Separation and capture of CO2 from large stationary sources and sequestration in 535

geological formations- coalbeds and deep saline aquifers,” J. Air Waste Manage. Assoc., 536

53(6), pp. 645–715. 537

[16] Granite, E. J., and O’Brien, T., 2005, “Review of novel methods for carbon dioxide 538

separation from flue and fuel gases,” Fuel Process. Technol., 86(14), pp. 1423–1434. 539

[17] Aaron, D., and Tsouris, C., 2005, “Separation of CO2 from flue gas: a review,” Sep. Sci. 540

Technol., 40(1-3), pp. 321–348. 541

[18] Manzolini, G., Campanari, S., Chiesa, P., Giannotti, A., Bedont, P., and Parodi, F., 2012, 542

“CO2 Separation From Combined Cycles Using Molten Carbonate Fuel Cells,” J. Fuel Cell 543

Sci. Technol., 9, p. 011018. 544

[19] Amorelli, A., Wilkinson, M. B., Bedont, P., Capobianco, P., Marcenaro, B., Parodi, F., and 545

Torazza, A., 2002, “An experimental investigation into the use of molten carbonate fuel 546

cells to capture CO2 from gas turbine exhaust gases,” Energy, 29, pp. 1279–1284. 547

[20] Winnick, J., Toghiani, H., and Quattrone, P. D., 1982, “Carbon dioxide concentration for 548

manned spacecraft using a molten carbonate electrochemical cell,” AIChE J., 28(1), pp. 549

103–111. 550

[21] Sugiura, K., Yanagida, M., Tanimoto, K., and Kojima, T., 2000, “The removal 551

characteristics of carbon dioxide in molten carbonate for the thermal power plant,” Proc. 552

GHGT-5, Aust. 553

[22] Rheinhardt, J., and Buttry, D. A., 2014, “Energy Efficient Capture and Release of Carbon 554

Dioxide in Tetraalkyl Phosphonium and Tetraalkyl Ammonium Ionic Liquids,” Meet. 555

Abstr., MA2014-02(25), p. 1431. 556


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


27

[23] Hasani, M., and Buttry, D., 2013, “Chemical Reactivity of Alkyl Thiolates Used in 557

Electrochemical CO2 Capture in Ionic Liquids,” 224th ECS Meeting, p. 2601. 558

[24] Buttry, D. A., 2014, “Capture and Release of Carbon Dioxide.” 559

[25] Li, K., and Li, N., 1993, “Removal of carbon dioxide from breathing gas mixtures using an 560

electrochemical membrane cell,” Sep. Sci. Technol., 28, pp. 1085–1090. 561

[26] Li, K., Teo, W., and Hughes, R., 1994, “Use of membranes for carbon dioxide removal in 562

underwater life support systems,” Underw. Technol., 20(1), pp. 13–17. 563

[27] Xiao, S., and Li, K., 1997, “On the use of an electrochemical membrane module for removal 564

of CO2 from a breathing gas mixture,” Chem. Eng. Res. Des., 75(4), pp. 438–446. 565

[28] Pennline, H. W., Granite, E. J., Luebke, D. R., Kitchin, J. R., Landon, J., and Weiland, L. 566

M., 2010, “Separation of CO2 from flue gas using electrochemical cells,” Fuel, 89(6), pp. 567

1307–1314. 568

[29] Landon, J., and Kitchin, J. R., 2010, “Electrochemical Concentration of Carbon Dioxide 569

from an Oxygen/Carbon Dioxide Containing Gas Stream,” J. Electrochem. Soc., 157(8), p. 570

B1149. 571

[30] Varcoe, J. R., Atanassov, P., Dekel, D. R., Herring, A. M., Hickner, M. A., Kohl, P. A., 572

Kucernak, A. R., Mustain, W. E., Nijmeijer, K., Scott, K., Xu, T., and Zhuang, L., 2014, 573

“Anion-exchange membranes in electrochemical energy systems,” Energy Environ. Sci., 574

7(10), pp. 3135–3191. 575

[31] Li, N., Leng, Y., Hickner, M. A., and Wang, C. Y., 2013, “Highly stable, anion conductive, 576

comb-shaped copolymers for alkaline fuel cells,” J. Am. Chem. Soc., 135(27), pp. 10124–577

10133. 578

[32] Hickner, M. A., Herring, A. M., and Coughlin, E. B., 2013, “Anion exchange membranes: 579


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


28

Current status and moving forward,” J. Polym. Sci. Part B Polym. Phys., 51(24), pp. 1727–580

1735. 581

[33] Grew, K. N., Ren, X., and Chu, D., 2011, “Effects of Temperature and Carbon Dioxide on 582

Anion Exchange Membrane Conductivity,” Electrochem. Solid-State Lett., 14(12), p. B127. 583

[34] Kiss, A. M., Myles, T. D., Grew, K. N., Peracchio, A. A., Nelson, G. J., and Chiu, W. K. 584

S., 2013, “Carbonate and Bicarbonate Ion Transport in Alkaline Anion Exchange 585

Membranes,” J. Electrochem. Soc., 160(9), pp. F994–F999. 586

[35] Myles, T. D., Kiss, A. M., Grew, K. N., Peracchio, A. A., Nelson, G. J., and Chiu, W. K. 587

S., 2011, “Calculation of Water Diffusion Coefficients in an Anion Exchange Membrane 588

Using a Water Permeation Technique,” J. Electrochem. Soc., 158(7), p. B790. 589

[36] Vega, J. A., Shrestha, S., Ignatowich, M., and Mustain, W. E., 2012, “Carbonate Selective 590

Ca2Ru2O7-y Pyrochlore Enabling Room Temperature Carbonate Fuel Cells,” J. 591

Electrochem. Soc., 159(1), p. B12. 592

[37] Vega, J. A., Spinner, N., Catanese, M., and Mustain, W. E., 2012, “Carbonate Selective 593

Ca2Ru2O7−y Pyrochlore Enabling Room Temperature Carbonate Fuel Cells,” J. 594

Electrochem. Soc., 159(1), p. B19. 595

[38] Hancock, C. A., Ong, A. L., and Varcoe, J. R., 2014, “Effect of carbonate anions on Bi-596

doped Ca2Ru2O7 pyrochlores that are potential cathode catalysts for low temperature 597

carbonate fuel cells,” RSC Adv., 4(57), p. 30035. 598

[39] Spinner, N., and Mustain, W. E., 2013, “Electrochemical Methane Activation and 599

Conversion to Oxygenates at Room Temperature,” J. Electrochem. Soc., 160(11), pp. 600

F1275–F1281. 601

[40] Spinner, N., and Mustain, W. E., 2011, “Effect of nickel oxide synthesis conditions on its 602


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


29

physical properties and electrocatalytic oxidation of methanol,” Electrochim. Acta, 56(16), 603

pp. 5656–5666. 604

[41] Gunasekara, I., Lee, M., Abbott, D., and Mukerjee, S., 2012, “Mass Transport and Oxygen 605

Reduction Kinetics at an Anion Exchange Membrane Interface: Microelectrode Studies on 606

Effect of Carbonate Exchange,” ECS Electrochem. Lett., 1(2), pp. F16–F19. 607

[42] Li, G., Wang, Y., Pan, J., Han, J., Liu, Q., Li, X., Li, P., Chen, C., Xiao, L., Lu, J., and 608

Zhuang, L., 2015, “Carbonation effects on the performance of alkaline polymer electrolyte 609

fuel cells,” Int. J. Hydrogen Energy, 40(20), pp. 6655–6660. 610

[43] Poynton, S. D., Slade, R. C. T., Omasta, T. J., Mustain, W. E., Escudero-Cid, R., Ocón, P., 611

and Varcoe, J. R., 2014, “Preparation of radiation-grafted powders for use as anion 612

exchange ionomers in alkaline polymer electrolyte fuel cells,” J. Mater. Chem. A, 2, p. 613

5124. 614

[44] Roen, L. M., Paik, C. H., and Jarvi, T. D., 2004, “Electrocatalytic Corrosion of Carbon 615

Support in PEMFC Cathodes,” Electrochem. Solid-State Lett., 7(1), p. A19. 616

[45] Shrestha, S., Liu, Y., and Mustain, W. E., 2011, “Electrocatalytic Activity and Stability of 617

Pt clusters on State-of-the-Art Supports: A Review,” Catal. Rev., 53(3), pp. 256–336. 618

[46] Vega, J. A., Chartier, C., and Mustain, W. E., 2010, “Effect of hydroxide and carbonate 619

alkaline media on anion exchange membranes,” J. Power Sources, 195(21), pp. 7176–7180. 620

[47] Wu, J., Yadav, R. M., Liu, M., Sharma, P. P., Tiwary, C. S., Ma, L., Zou, X., Zhou, X., 621

Yakobson, B. I., Lou, J., and Ajayan, P. M., 2015, “Achieving Highly Efficient, Selective, 622

and Stable CO2 Reduction on Nitrogen-Doped Carbon Nanotubes,” Nano, 9(5), pp. 5364–623

5371. 624

625


Accep

ted

Manus

crip

t Not

Cop

yedi

ted


c accepted manuscript not copyedited - university of surreyepubs.surrey.ac.uk/811908/1/jeecs-16-1019...

Documents